source: https://doi.org/10.7892/boris.127193 | downloaded: 11.3.2021

Accepted Manuscript

Revisiting systemic treatment of bacterial endophthalmitis: a review of intravitrealpenetration of systemic antibiotics

Lisa Brockhaus, David Goldblum, Laura Eggenschwiler, Stefan Zimmerli, CatiaMarzolini

PII: S1198-743X(19)30039-4

DOI: https://doi.org/10.1016/j.cmi.2019.01.017

Reference: CMI 1563

To appear in: Clinical Microbiology and Infection

Received Date: 14 December 2018

Accepted Date: 28 January 2019

Please cite this article as: Brockhaus L, Goldblum D, Eggenschwiler L, Zimmerli S, Marzolini C,Revisiting systemic treatment of bacterial endophthalmitis: a review of intravitreal penetration ofsystemic antibiotics, Clinical Microbiology and Infection, https://doi.org/10.1016/j.cmi.2019.01.017.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service toour customers we are providing this early version of the manuscript. The manuscript will undergocopyediting, typesetting, and review of the resulting proof before it is published in its final form. Pleasenote that during the production process errors may be discovered which could affect the content, and alllegal disclaimers that apply to the journal pertain.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Revisiting systemic treatment of bacterial endophthalmitis: a review of intravitreal

penetration of systemic antibiotics

Lisa Brockhaus1, David Goldblum2, Laura Eggenschwiler2, Stefan Zimmerli3, Catia Marzolini1 1 Division of Infectious Diseases and Hospital Epidemiology, Departments of Medicine and

Clinical Research, University Hospital of Basel and University of Basel, Basel, Switzerland.

2 Department of Ophthalmology, University Hospital of Basel and University of Basel,

Switzerland.

3 Department of Infectious Diseases, University Hospital of Bern and University of Bern,

Switzerland

Narrative review

Word count: Abstract: 221

Text: 2497

Tables: 2

Corresponding author:

Lisa Brockhaus, MD

Division of Infectious Diseases

University Hospital Basel

Switzerland

Phone: +41 61 265 50 53

E-mail: lisa.brockhaus@unibas.ch

Keywords

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Endophthalmitis, systemic antibiotic, vitreous concentration, intravitreal penetration, blood-

retinal barrier

Abstract

Background: Adjunctive systemic antibiotic therapy for treatment of bacterial

endophthalmitis is controversial but common practice due to the severity of the disease. In

absence of guidance documents, several antibiotic regimens are being used without applying

evidence-based prescribing, thus leading to inappropriate treatment of this serious eye

condition.

Objectives: To summarize available data on intravitreal penetration of systemically

administered antibiotics and to discuss their usefulness from a microbiological and

pharmacological point of view.

Sources: We performed a systematic PubMed search of studies investigating antibiotic

concentrations in the vitreous after systemic administration in humans, and selected animal

models.

Content: The best-documented agents achieving therapeutic levels in the vitreous are

meropenem, linezolid and moxifloxacin. Vancomycin, cefazoline, ceftriaxone, ceftazidime,

imipenem and trimethoprim-sulfamethoxazole reach levels justifying their use in specific

situations. Available data do not support the use of ciprofloxacin, levofloxacin,

aminoglycosides, aminopenicillins, piperacillin, cefepime, and clarithromycin. With very

limited but available promising data, the use of daptomycin and rifampicin deserves further

investigation.

Implications: The choice of the adjunctive systemic antibiotic agent – in situations where

considered relevant for treatment - must to date be made on an individual base, considering

microbiological aspects as well as operative status and inflammation of the eye. This review

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

gives a systematic overview of antibiotic options and provides guidance to the clinician

striving for optimal systemic antibiotic treatment of bacterial endophthalmitis.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Introduction

Vitrectomy and intravitreal antibiotics are nowadays considered as the gold standard

treatment of bacterial endophthalmitis [1]. Adjunctive use of systemic antibiotics is

controversial but common practice justified by the severity of the disease. Data on intravitreal

antibiotic levels reached by systemic administration are sparse and comprehensive

recommendations for systemic use have not been established [1–3]. The availability of new

antibiotic agents, changing resistance patterns, and new surgical techniques using implants

such as keratoprosthesis justify revisiting their systemic use for endophthalmitis.

This review summarizes available data on intravitreal penetration of systemically

administered antibiotics and discusses preferred regimens for the treatment of bacterial

endophthalmitis from a microbiological and pharmacological perspective.

Definition and commonly isolated microorganisms

Endophthalmitis refers to the inflammation of the internal eye affecting the vitreous cavity

and the anterior chamber, resulting from exogenous (mostly surgery related) or, more rarely

(5-10%) hematogenous insertion of microorganisms [4].

The most commonly isolated microorganisms are coagulase-negative staphylococci (40-

70%), Staphylococcus aureus (10-17%), streptococci (5-15%), other Gram-positive cocci

including enterococci (5%), as well as Gram-negative bacilli (5-10%) including Haemophilus

influenzae and Pseudomonas aeruginosa [5–8]. The microbiologic spectrum of less common

forms like post-traumatic or endogenous endophthalmitis is more varied. For instance,

bacillus sp. is regularly found after open-globe injuries [9]. The bacterial spectrum

encountered may further vary according to local epidemiology and peri-operative

prophylactic regimens [10].

Current treatment recommendations

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

The current mainstays of bacterial endophthalmitis treatment are vitrectomy and intravitreal

antibiotics [1]. Vitrectomy aims to reduce the bacterial load with the intention of a local

“source control”. Although complete vitrectomy would ensure maximal eradication of the

infected tissue, partial vitrectomy is often preferred in clinical practice, because of the lower

associated risk of iatrogenic retinal detachment.

Intravitreal antibiotics are injected immediately after vitrectomy and their administration is

usually repeated after 48 hours if the clinical course is not favourable. The most commonly

used regimens are vancomycin combined with ceftazidime or amikacin. The downside of

repeated injections is an increased risk of retinal toxicity.

Intravitreal dexamethasone is often added to reduce intraocular inflammation despite

conflicting evidence [1]. Corticosteroids accelerate blood-retinal barrier restitution and thus

influence antibiotic penetration.

The role of systemic antibiotics

The single large clinical trial evaluating the role of systemic antibiotics for the treatment of

endophthalmitis is the Endophthalmitis-Vitrectomy-Study conducted in the 1990s [11]. In

this randomized study, no significant difference in visual acuity was found in patients

receiving intravitreal antibiotics followed by intravenous antibiotic therapy compared to

patients receiving only intravitreal treatment. However, this finding has been questioned

because of the exclusion of patients with severe endophthalmitis and the choice of adjunctive

antibiotics: ceftazidime has poor activity against the dominant Gram-positive organisms and

amikacin has very limited intraocular penetration. Since then, only small studies have

evaluated the efficacy of systemic antibiotic therapy in endophthalmitis with varying

methodologies and results [12-13].

Some recommendations advocate the adjunctive use of systemic antibiotics in severe acute

purulent postoperative endophthalmitis [1]. Recommended regimens include vancomycin

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

combined with ceftazidime [14] or imipenem with ciprofloxacin [15]. For the treatment of

endogenous endophthalmitis the use of systemic antibiotics is undisputed [2, 4] .

The pharmacokinetic rationale for adjunctive systemic antibiotics is the rapid elimination of

intravitreally applied antibiotics, with almost complete removal after 24h [16], whereas

systemic administration favors intraocular antibiotic accumulation over time.

Discussing the controversial benefit of adjunctive systemic antibiotics in terms of visual

outcome is beyond the scope of this review. The imminent poor outcome constitutes a strong

argument for clinicians to use systemic therapy. We strongly believe that adjunctive systemic

therapy should not be denied on an individual base, provided that all efforts are made to

isolate the causative pathogen and to apply evidence-based prescribing of antibiotic agents.

Intravitreal penetration of systemic antibiotics

Factors determining the penetration of antibiotics into the eye

The penetration of antibiotics into the posterior segment of the eye after systemic

administration is limited by two blood-retinal barrier mechanisms (BRB): The retinal

pigment endothelial cells located within the retinal cell layers (outer BRB) and the retinal

capillary endothelial cells (inner BRB) [17]. Of note, entry into the anterior segment of the

eye is limited by the blood-aqueous-barrier characterized by less restrictive properties,

thereby resulting in different aqueous and vitreous drug concentrations.

Drug permeability across the blood-retinal barriers depends on drug characteristics such as

the molecular size, lipophilicity, ionization and protein binding. Of interest, BRB was shown

to be more permeable than the blood-brain-barrier (BBB) owing to morphological differences

[17]. As in the BBB, ocular inflammation increases drug permeability across the BRB.

Elimination of antibiotics from the vitreous occurs via two routes: passive diffusion to the

anterior chamber and through the Schlemm’s channel (anterior route), and retrograde

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

transport through the blood-retinal barrier (posterior route). Clearance pathways from the

vitreous depends not only on physico-chemical drug properties and ocular inflammation, but

also on the surgical status [18]. Post-operative aphakia – after removal of an artificial intra-

ocular lens – as well as vitrectomy influence elimination of antibiotics [19].

Pharmacokinetic studies

The knowledge of antibiotics pharmacokinetics in the eye is derived from two types of

studies: single concentration measurements in human eyes performed at the time of surgery,

and rabbit models allowing for repetitive drug measurements. Several limitations should be

considered: although single drug measurements are of high value for antibiotics characterized

by a concentration-dependent killing effect (aminoglycosides, daptomycin), this approach is

less informative for antibiotics with a killing profile that depends on time-above-MIC (beta-

lactams) or AUC/MIC (fluoroquinolones, vancomycin, linezolid).

1) Common pharmacodynamic index values (such as Cmax/MIC used in most of the

assessed studies) do not necessarily reflect efficacy in the complex microenvironment

of the eye.

2) Comprehensive MIC-studies of endophthalmitis isolates are missing. In this review

we use EUCAST breakpoints and wild-type distributions (ECOFF) [20] (table 2) as a

reference to estimate potential efficacies of antibiotics.

3) Animal studies must be interpreted with caution, as they do not fully reflect the

pharmacokinetics in humans.

Suitable publications were identified by a PubMed search using the terms [antibiotic name]

and [vitreous] and [systemic/oral/intravenous]. If no publications were found, the search was

complemented by the terms [eye] and [penetration]. Rabbit model studies were included for

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

antibiotics with limited human data or if they provided additional relevant information.

Studies are summarized in table 1.

Review of available literature

Vancomycin

Vancomycin did not show any accumulation in phakic (including inflamed) eyes in a rabbit

study [21]. Aphakic-vitrectomized eyes showed vitreous levels just above breakpoints of

commonly involved Gram-positive organisms after one dose. In aphakic eyes, comparable

levels were only reached after prolonged therapy. The poor intravitreal penetration of

vancomycin is consistent with the limited CSF penetration [22] and explained by its high

molecular weight and hydrophilicity.

The rationale for its continued empiric use [14] despite limited data, is its microbiological

spectrum covering almost 100% of Gram-positive organisms causing endophthalmitis [7].

Available data nevertheless suggest that systemic administration should only be considered in

aphakic eyes. Given the delay in achieving sufficient concentrations, systemic administration

should follow immediately intravitreal injection of the same agent.

Penicillins

Poor vitreal penetration of ampicillin and amoxicillin were demonstrated in rabbit models

[23-24]. Similarly, insufficient vitreous concentrations of piperacillin were demonstrated in

human eyes with diverse operative status [25]. The vitreal penetration of penicillin G has not

been studied but based on CSF penetration data [22] and drug properties, penetration into the

vitreous is anticipated to be minimal. Nevertheless, in analogy to CSF infections, high

systemic doses might provide effective vitreous levels for streptococcal endophthalmitis.

Available data suggest that systemic administration of most penicillins is not appropriate to

treat endophthalmitis. Penicillin G vitreous levels remain to be investigated.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Cephalosporins

Cefazoline vitreous concentrations are issued from a large rabbit study [26] showing levels

well above streptococcal breakpoints in aphakic-vitrectomized inflamed eyes but

undetectable levels in phakic non-inflamed eyes.

Ceftriaxone was detectable in human phakic non-inflamed eyes after multiple dosing [27] at

levels well above streptococci and enterobacteriaceae breakpoints, but below the ECOFF of

S.aureus. Whether the observed levels result from accumulation after repetitive dosing, or

rapid penetration as previously shown in CSF studies [28], is not known.

Vitreous levels of ceftazidime are based on two rabbit studies [29-30]. Levels were above

enterobacteriaceae breakpoints in aphakic-vitrectomized inflamed eyes. Only delayed

penetration was observed in non-vitrectomized inflamed eyes and undetectable levels were

found in phakic non-inflamed eyes [29].

Cefepime showed low vitreous levels in human phakic non-inflamed eyes [31], nonetheless

penetration in inflamed eyes have not been studied.

In summary, cefazoline, ceftriaxone and ceftazidime can be considered as targeted therapy

when the pathogen is identified and the MIC of the isolate has been determined, yet on a thin

evidence base. Ceftazidime is likely to be effective against most enterobacteriaceae in

aphakic-vitrectomized eyes. Since it exhibits very limited anti-streptococcal and no anti-

staphylococcal activity, it should not be used to cover Gram-positive organisms. Cefepime

cannot be recommended as observed levels are clearly below those of other cephalosporins,

yet investigation in inflamed eyes is warranted. Neither ceftazidime nor cefepime can be

recommended to treat pseudomonas endophthalmitis based on available data.

Carbapenems

Two studies in human phakic non-inflamed eyes showed imipenem vitreous concentrations

around 2.0 ug/ml [32-33]. Meropenem was shown to rapidly achieve four-fold higher

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

vitreous concentrations in a comparable study [34]. This finding is consistent with observed

high meropenem levels in CSF [22] and explained by favorable physicochemical properties.

Unlike imipenem, the observed levels of meropenem clearly exceed breakpoints for relevant

Gram-positive and Gram-negative organisms.

The potential value of carbapenems in empirical treatment – as single agents - is limited by

the high prevalence of oxacillin-resistent coagulase negative staphylococci also in ophthalmic

isolates [8]. For empiric combination therapy, we would advocate the use of meropenem

rather than imipenem based on the above-discussed data. Furthermore, meropenem appears to

be the preferred option for targeted Pseudomonas treatment, particularly when considering

the insufficient concentrations of ciprofloxacin, piperacillin, ceftazidime and cefepime as

discussed above.

Rifampicin

High rifampicin vitreous levels were observed in phakic non-inflamed rabbit eyes [35],

however with doses that, corrected for weight, would largely exceed tolerated doses in

humans.

Human studies are limited to aqueous levels, observed to be 0.2-1.3 ug/ml [36]. As

rifampicin vitreous levels were consistently shown to be half of aqueous levels [31], human

vitreous levels could be expected to exceed 0.1 ug/ml, which is well above breakpoints for

staphylococci.

A role of rifampicin in endophthalmitis after foreign-body implantation, incomplete

vitrectomy, and aggressive S.aureus-infection deserves further consideration due to its

bactericidal and biofilm-active-properties. Importantly, rifampicin has to be combined with

an effective second anti-staphylococcal agent to prevent resistance.

Linezolid

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Mean vitreous levels of linezolid in phakic non-inflamed human eyes range from 1.2 to 3.7

ug/ml after one dose [37–40], with two studies showing further accumulation after two doses

(4.5 and 5.7 ug/ml) [37, 39]. The good penetration into the vitreous is consistent with CSF

penetration [22].

Considering breakpoints of Gram-positive organisms, sufficient vitreous levels are likely to

be attained after two doses. The drug, however, is bacteriostatic.

The well documented ocular penetration and comprehensive coverage of Gram-positive

germs make of linezolid a potential alternative to vancomycin. Caution is needed due to its

toxicity (myelosuppression, peripheral neuropathy, optic neuropathy), although relatively

infrequent in short-term administration [41].

Daptomycin

Knowledge of daptomycin penetration is limited to one case report of a patient treated for

MRSA-endophthalmitis in a strongly inflamed, phakic, non-vitrectomized eye [42]. Single

dose administration (10 mg/kg) resulted in a vitreous concentration of 12.4 ug/ml 42h post

administration (patient had renal insufficiency). This finding appears promising considering

low staphylococcal breakpoints and the drug’s bactericidal properties, but its use remains

experimental to date.

Aminoglycosides

Vitreous concentrations far below breakpoints of relevant pathogens were shown after

intravenous administration of amikacin and gentamycin in a rabbit study [43], despite study

conditions expected to enhance vitreous levels (inflammation, aphakia and vitrectomy).

Given the availability of better alternatives, there is no role for systemic administration of

aminoglycosides in endophthalmitis.

Fluoroquinolones

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Despite relatively good vitreous/serum (V/S) ratios owing to favourable physicochemical

properties, observed concentrations in vitreous proved insufficient for ciprofloxacin [44–50]

and levofloxacin [38, 51-52].

Moxifloxacin demonstrated concentrations well above breakpoints of relevant organisms

after two doses [53-54], whereas concentrations were significantly lower after a single dose

[55-56]. The low concentrations of the Vedantham study can be explained by an inadequate

short sampling time of 90 minutes after oral administration [56] . Moxifloxacin maximal CSF

levels were shown to occur 2-3 hours after maximal systemic levels [57].

In summary, there are sufficient data to oppose the use of ciprofloxacin and levofloxacin for

the treatment of endophthalmitis. Conversely, several studies consistently demonstrating

satisfactory moxifloxacin levels are available. Given the observed higher levels with

moxifloxacin 800mg, this increased dosage can be considered with a careful monitoring for

side effects. Safety data on this dosage are not comprehensive to date [58].

Moxifloxacin lacks activity against most oxacillin-resistant coagulase-negative staphylococci

[8]), and streptococci rapidly develop resistance particularly when drug concentrations are

low. Therefore, the use of moxifloxacin for empirical treatment of endophthalmitis is limited.

Trimethoprim-Sulfamethoxazole

Sulfonamides and trimethoprim have demonstrated a moderate penetration in the vitreous in

one human study [59]. The doses used, however, were below those for the treatment of

meningitis [60]. The observed concentrations are not exceeding breakpoints of all relevant

organisms although higher concentrations might be reached with increased doses.

Concentrations are far below the breakpoint of Stenotrophomonas maltophilia (4 ug/ml)

suggesting difficulty in treating this pathogen.

Due to limited data and better alternatives for Gram-positive organisms, there is no current

role for TMP/SMX in systemic endophthalmitis treatment.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Clarithromycin

Based on one human study of phakic non-inflamed eyes showing insufficient vitreous levels

[61], there is no argument to recommend clarithromycin for endophthalmitis treatment.

Similarly insufficient levels have been reported in CSF, where the use of clarithromycin is

limited to case reports of successful treatment of atypic organisms [22].

Conclusion

Data on the intravitreal concentrations of systemic antibiotics are generally scarce and are

based on a single study for many agents. Relatively good evidence exists for therapeutic

vitreous levels of meropenem, linezolid and high-dose moxifloxacin. None covers the

required bacterial spectrum when used empirically as single agents, but the combination of

linezolid with meropenem in empirical treatment of endophthalmitis may offer broad activity

against the majority of pathogens. Vancomycin, cefazoline, ceftriaxone, ceftazidime,

imipenem, daptomycin and TMP-SMX exhibit levels supporting their use in specific

situations for targeted therapy. The operative status of the infected eye needs also to be

considered. Available data do not support the use of ciprofloxacin, levofloxacin,

aminoglycosides, aminopenicillins, piperacillin, cefepime or clarithromycin. Rifampicin may

be considered for combination therapy in complicated staphylococcal infections. Further data

on daptomycin vitreous penetration would be valuable.

The choice of the adjunctive systemic antibiotic agent – in situations where considered

relevant for treatment - must to date be made on an individual base, taking into account

suspected or detected organisms, operative status, intraocular inflammatory activity and drug

side effect profiles. Future research should assess the clinical outcome after use of systemic

antibiotics with documented good intraocular penetration (e.g. meropenem, linezolid and

moxifloxacin), and assess the role of rifampicin for staphylococcal infections.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Transparency declaration

The authors declare no competing interests regarding this manuscript. This work did not

receive any funding.

Access to data

All studies included in this review are publicly available.

Author contributions

LB has done the literature search, analyzed and compiled the data. DG, LE have provided

input in relation to ophthalmological and ophthalmo-surgical aspects. CM has contributed to

the pharmacological content and analysis of data. SZ has participated in the interpretation of

data and revised the manuscript. LB and CM have drafted the manuscript. All authors have

revised the manuscript for intellectual content and approved the final submitted version.

Acknowledgement

We thank Prof. Manuel Battegay for helpful comments on an earlier draft of this paper.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

References

[1] Barry P, Cordovés L, Gardner S. ESCRS Guidelines for Prevention and Treatment of Endophthalmitis Following Cataract Surgery: Data, Dilemmas and Conclusions. 2013. http://www.escrs.org/endophthalmitis

[2] Birnbaum F, Gupta G. Endogenous Endophthalmitis: Diagnosis and Treatment. EyeNet Mag 2016;6:33–35.

[3] Wakely L, Sheard R, Hospital RH. Recent advances in Endophthalmitis Management. Focus 2014;2:5-6.

[4] Kernt M, Kampik A. Endophthalmitis : Pathogenesis, clinical presentation, management, and perspectives. Clin Ophthalmol 2010;4:121–135.

[5] Ong A, Te NA, Zagora S, Symes R, Yates W, Chang A et al. Post-surgical versus post-intravitreal injection endophthalmitis: changing patterns in causative flora. Clin Exp Ophthalmol 2018;jul:1–6. (Epub ahead of print)

[6] Han DP, Wisniewski SR, Wilson LA, Barza M, Vine AK, Doft BH et al. Spectrum and susceptibilities of microbiologic isolates in the endophthalmitis vitrectomy study. Am J Ophthalmol 1996;122(1),1–17.

[7] Gentile RC, Shukla S, Shah M, Ritterband DC, Engelbert M, Davis A et al. Microbiological spectrum and antibiotic sensitivity in endophthalmitis: A 25-year review. Ophthalmology 2014;121(8):1634–1642.

[8] Asbell PA, Mah FS, Sanfilippo CM, DeCory HH. Antibiotic susceptibility of bacterial pathogens isolated from the aqueous and vitreous humor in the Antibiotic Resistance Monitoring in Ocular Microorganisms (ARMOR) surveillance study. J Cataract Refract Surg 2016;42(12):1841–1843.

[9] Miller JJ, Scott IU, Flynn HW jr, Smiddy WE, Murray TG, Berrocal A et al. Endophthalmitis Caused by Bacillus Species. Am J Ophthalmol 2008;145(5):883–888.

[10] Friling E, Lundström M, Stenevi U, Montan P. Six-year incidence of endophthalmitis after catarcat surgery: Swedish national study. J Cataract Refract Surg 2013;39(1):15–21.

[11] Endophthalmitis Vitrectomy Study Group. Results of the Endophthalmitis Vitrectomy Study. A randomized trial of immediate vitrectomy and of intravenous antibiotics for the treatment of postoperative bacterial endophthalmitis. Arch Ophthalmol 1995;113(12):1479–1496.

[12] Tappeiner C, Schuerch K, Goldblum D, Zimmerli S, Fleischhauer JC, Frueh BE. Combined meropenem and linezolid as a systemic treatment for postoperative endophthalmitis. Klin Monatsbl Augenheilkd 2010;227(4):257–261.

[13] Hooper CY, Lightman SL, Pacheco P, Tam PMK, Khan A, Taylor SRJ. Adjunctive antibiotics in the treatment of acute bacterial endophthalmitis following cataract surgery. Acta Ophthalmol 2012;90(7):572–573.

[14] Behrens-Baumann W. Current Therapy for Postoperative Endophthalmitis. Klin Monatsbl für Augenheilk 2008;225(11):919–923.

[15] German Society for Cataract & Refractive Surgeons. Leitlinie zur Prophylaxe und Therapie von Endophthalmitiden. 2005.

http://www.dgii.org/de/empfehlungen-101.html

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

[16] Ficker L, Meredith T, Gardner S, Wilson LA. Cefazolin levels after intravitreal injection. Effects of inflammation and surgery. Invest Ophthalmol Vis Sci 1990;31(3):502–505.

[17] Del Amo E, Rimpelä AK, Heikkinen E, Kari OK, Ramsay E, Lajunen T, et al. Progress in Retinal and Eye Research Pharmacokinetic aspects of retinal drug delivery. Prog Retin Eye Res 2017;57:134–185.

[18] Luaces-Rodríguez A, González-Barcia M, Blanco-Teijeiro M, Gil-Martinez M, Gonzalez F, Gomez-Ulla F et al. Review of Introcular Pharmakokinetics of Anti-Infectives Commonly Used in the Treatment of Infectious Endophthalmitis. Pharmaceutics 2018;10(66);1–15.

[19] Radhika M, Mithal K, Bawdekar A, Dave V, Jindal A, Relhan N, et al. Pharmacokinetics of intravitreal antibiotics in endophthalmitis. J Ophthalmic Inflamm Infect 2014;4(22):1–9.

[20] European Comitee for Suscepitibility Testing.

http://www.eucast.org/clinical_breakpoints. Version 8.1 (accessed in May 2018)

[21] Meredith TA, Aguilar HE, Shaarawy A, Kincaid M, Dick J, Niesman MR. Vancomycin levels in the vitreous cavity after intravenous administration. Am J Ophthalmol 1995;119(6):774–778.

[22] Nau R, Sörgel F, Eiffert H. Penetration of drugs through the blood-cerebrospinal fluid/blood-brain barrier for treatment of central nervous system infections. Clin Microbiol Rev 2010;23(4):858–883.

[23] Salminen L. Ampicillin penetration into the rabbit eye. Acta Ophthalmol 1878;56: 977–938.

[24] Faigenbaum S, Boyle G, Prywes A, Abel R, Leopold I. Intraocular Penetration of Amoxicillin. Am J Ophthalmol 1976;82(4):598–603.

[25] Robinet A, Le Bot M, Colin J, Riche C. Penetration of Piperacillin into the vitreous after Intravenous Administration. Retina 1998;18(6):526–530.

[26] Martin DF, Ficker LA, Gardner SK, Louis A, Doses RA. Vitreous cefazolin levels after intravenous injection. Arch Ophthalmol 1990;108: 411–414.

[27] Sharir M, Triesrer G, Kneer J, Rubinstein E. The Intravitreal Penetration of Ceftriaxone in Man following Systemic Administration. Invest Ophthalmol Vis Sci 1989;30(10):2179–2183.

[28] Del Rio M, McCracken G, Nelson J, Chrane D, Shelton S. Pharmacokinetics and cerebrospinal fluid bactericidal activity of ceftriaxone in the treatment of pediatric patients with bacterial meningitis. Antimicrob Agents Chemother 1982;22(4):622–627.

[29] Aguilar H, Meredith T, Shaarawy A, Kincaid M, Dick J. Vitreous cavity penetration of ceftazidime after intravenous administration. Retina 1995;15:154–159.

[30] Walstad R, Blika S. Penetration of Ceftazidime into the Normal Rabbit and Human Eye. Scand J Infect Dis 1985;44:63–67.

[31] Aras C, Ozdamar A, Ozturk R, Karacorla M, Ozkan S. Intravitreal penetration of cefepime after systemic administration to humans. Ophthalmologica 2002;216:261–264.

[32] Adenis J, Mounier M, Salomon J, Denis F. Human vitreous penetration of imipenem.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Eur J Ophthalmol 1994;4(2):115–117.

[33] Axelrod J, Newton J, Klein R, Bergen R, Sheikh M. Penetration of Imipenem Into Human Aqueous and Vitreous Humor. Am J Ophthalmol 1987;104:649–653.

[34] Schauersberger J, Amon M, Wedrich A, Nepp J, El Menyawi I, Derbolav A, et al. Penetration and decay of meropenem into the human aqueous humor and vitreous. J Ocul Pharmacol Ther 1999;15(5):439–445.

[35] Wong K, D’Amico D, Oum B, Baker P, Kenyon K. Intraocular penetration of rifampin after oral administration. Graefe’s Arch Exp Ophthalmol 1990;228:40–43.

[36] Outman WR, Levitz RE, Hill DA, Nightingale CH. Intraocular penetration of rifampin in humans. Antimicrob. Agents Chemother 1992;36(7):1575–1576.

[37] Fiscella RG, Lai W, Buerk B, Khan M, Rodvold KA, Pulido JS, et al. Aqueous and vitreous penetration of linezolid after oral administration. Ophthalmology 2004;111(6):1191–1195.

[38] George J, Fiscella R, Blair M, Rodvold K, Ulanski L, Stokes J, et al. Aqueous and vitreous penetration of linezolid and levofloxacin after oral administration. J Ocul Pharmacol Ther 2010;26(6):579–586.

[39] Horcajada JP, Atienza R, Sarasa M, Soy D, Adán A, Mensa J. Pharmacokinetics of linezolid in human non-inflamed vitreous after systemic administration. J Antimicrob Chemother, vol. 63, no. 3, pp. 550–552, 2009.

[40] Ciulla T, Comer G, Peloquin C, Wheeler J. Human vitreous distribution of linezolid after a single oral dose. Retina 2005;25:619–624.

[41] Vazquez J, Arnold AC, Swanson RN, Biswas P, Bassetti M. Safety of long-term use of linezolid : results of an open-label study. Ther Clin Risk Manag 2016;12:1347–1354.

[42] Sheridan K, Potoski B, Shields R, Nau G. Presence of Adequate Intravitreal Concentration of Daptomycin After Systemic Intravenous Administration in an Patient with Endogenous Endophthalmitis. Pharmacotherapy 2010;30:1247–1251.

[43] El-Massry A, Meredith TA, Aguilar HE; Shaarawy A, Kincaid M, Dick J, et al. Aminoglycoside levels in the rabbit vitreous cavity after intravenous administration. Am J Ophthalmol 1996;122(5):684–689.

[44] Morlet N, Graham GG, Gatus B, McLachlan AJ, Salonikas C, Naydoo D, et al. Pharmacokinetics of Ciprofloxacin in the Human Eye: a Clinical Study and Population Pharmacokinetic Analysis. Antimicrob Agents Chemother 2000;44(6):1674–1679.

[45] Cekic O, Batman C, Yasar Ü, Basci N, Bozkurt A, Kayaalp S. Human aqueous and vitreous humour levels of ciprofloxacin following oral and topical administration. Eye, 1999;13:555–558.

[46] El Baba F, Trousdale M, Gauderman W, Wagner D, Liggett P. Intravitreal penetration of oral ciprofloxacin in humans. Ophthalmology 1992;99(4):483–486.

[47] Keren G, Alhalel A, Bartov E, Kitzes-Cohen R, Rubinstein E, Segev S, et al. The intravitreal penetration of orally administered ciprofloxacin in humans. Invest Ophthalmol Vis Sci 1991;32(8):2388–2392.

[48] Sweeney G, Fern A, Lindsay G, Doig M. Penetration of ciprofloxacin into the aqueous humour of the uninflamed human eye after oral administration. J Antimicrob Chemother 1990;26(1):99–105.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

[49] Mounier M, Adenis J, Denis F. Intraocular penetration of ciprofloxacin after infusion and oral administration. Pathol Biol 1988;36(5):724–727.

[50] Madu AA, Mayers M, Perkins R, Liu W, Drusano GL, Aswani R, et al. Aqueous and vitreous penetration of ciprofloxacin following different modes of systemic administration. Exp Eye Res 1996;63(2):129–136.

[51] Fiscella R, Nguyen TKP, Cwik MJ, Philpotts BA, Friedlander D. Alter M et al. Aqueous and vitreous penetration of levofloxacin after oral administration. Ophthalmology 1999;106:2286–2290.

[52] Herbert EN, Pearce IA, McGalliard J, Wong D, Groenewald C. Vitreous penetration of levofloxacin in the uninflamed phakic human eye. Br J Ophthalmol 2002;86(4):387–389.

[53] Hariprasad S, Shah GK, Mieler WF, Feiner L, Blinder KJ, Holekamp NM, et al. Vitreous and Aqueous Penetration of Orally Administered Moxifloxacin in Humans. Arch Ophthalmol 2006;124:178–182.

[54] Fuller J, Lott MN, Henson N, Bhatti A, Singh H, McGwin G, et al.Vitreal penetration of oral and topical moxifloxacin in humans. Am J Ophthalmol 2007;143(2):338–340.

[55] Lott MN, Fuller J, Hancock H, Singh J, Singh H, McGwin G, et al.Vitreal penetration of oral moxifloxacin in humans. Retina 2008;28(3):473–476.

[56] Vedantham V, Lalitha P, Velpandian T, Ghose S, Mahalaksmi R, Ramasamy K. Vitreous and aqueous penetration of orally administered moxifloxacin in humans. Eye 2006;20:1273–1278.

[57] Alffenaar JWC, van Altena R, Bökkerink HJ, Luijckx GJ, van Soolingen D, Aarnoutse RE, et al. Pharmacokinetics of Moxifloxacin in Cerebrospinal Fluid and Plasma in Patients with Tuberculous Meningitis. Clin Infect Dis 2009;49(7):1080–1082.

[58] Ruslami R, Rizal Ganiem A, Dian S, Apriani L, Achmad TH, van der Ven AJ, et al. Intensified regimen containing rifampicin and moxifloxacin for tuberculous meningitis: An open-label, randomised controlled phase 2 trial. Lancet Infect Dis 2013;13(1):27–35.

[59] Feiz V, Nijm L, Glickman R, Morse L, Telander D, Park S, et al. Vitreous and aqueous penetration of orally administered trimethoprim-sulfamethoxazole combination in humans. Cornea 2013;2(10):1315–1320.

[60] Goodwin C, Bucens M, Davis R, Norcott T. High-dose co-trimoxazole and its penetration through uninflamed meninges. Med J Aust 1981;2(1):24–27.

[61] Al-Sibai MB, Al-Kaff AS, Raines D, El-Yazigi A. Ocular penetration of oral clarithromycin in humans. J Ocul Pharmacol Ther 1998;14(6)575–583.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

table 1. Studies evaluating vitreous concentrations after systemic administration of antibiotics

antibiotic model operative statusa

inflammation subject dose mean Cmax after 1 dose V/S vitreous antibiotic references

statusb

number (ug/ml) level > MICc

Vancomycin rabbit ph/aph/a-v inf+non-inf 58 15 mg/kg iv 5.4 (a-v inf), 1.1 (aph inf), 0.0 (ph inf) yes in a-v inf Meredith 1994 [21]

Ampicillin rabbit ph non-inf 49 50 mg/kg iv 0.1 no Salminen 1978 [23]

Amoxicillin rabbit ph non-inf 52 50/500 mg/kg iv (!) 0.2/1.6 0.02 Faigenbaum [24]

Piperacillin human ph/ps/aph inf+non-inf 45 4 g iv undetectable no Robinet 1998 [25]

Cefazoline rabbit ph/a-v inf+non-inf 40 50 mg/kg iv 6.7 (a-v inf), 3.0 (ph inf) yes in a-v inf Martin 1990 [26]

Ceftriaxon human ph non-inf 17 2g im bid 5.1 after 6 doses 0.04 partially Sharir 1998 [27]

Ceftazidime rabbit ph/aph/a-v inf+non-inf 46 50 mg/kg iv 35.4 (a-v inf), 0 (aph inf, ph inf) 0.3 partially Aguilar 1995 [29]

Ceftazidime rabbit ph non-inf 15 50 mg/kg iv <1 no Walstad 1985 [30]

Cefepime human ph non-inf 30 1g/2g iv 1.9/2.9 0.08 no Aras 2002 [31]

Imipenem human ph/ps non-inf 10 0.5g/1g iv 0.2/1.1 0.08 partially Adenis 1994 [32]

Imipenem human ph non-inf 10 1g iv 2.5 0.1 partially Axelrod 1987 [33]

Meropenem human ph non-inf 14 2g iv 8.9 0.3 yes Schauersberger 1999 [34]

Rifampicin rabbit ph non-inf ? 150/300/600mg po (!) 2.2/2.6/15.2 yes Wong 1990 [35]

Linezolid human ph non-inf 29 600mg po 2.3 0.3 partially Fiscella 2004 [37]

Linezolid human ph non-inf 16 600mg po 3.7 0.8 yes George 2010 [38]

Linezolid human ph non-inf 24 600mg iv/po 3.7 0.6 yes Horcajada 2008 [39]

Linezolid human ph non-inf 12 600mg po 1.2 0.1 partially Ciulla 2005 [40]

Daptomycin human ph inf 1 10 mg/kg iv 12.4 0.3 yes Sheridan 2010 [42]

Amikacin rabbit a-v inf 7 6 mg/kg iv 8.5 no El-Massry 1996 [43]

Gentamycin rabbit a-v inf 7 1.6 mg/kg iv 1.8 no El-Massry 1996 [43]

Ciprofloxacin human ph non-inf various 0.1-0.5 no Morlet 2000 [44], et al. [45-50]

Levofloxacin human ph non-inf 45 500mg po od/bid 0.6; 2.5 after 2 doses 0.3 no Fiscella 1999 [51]

Levofloxacin human ph non-inf 10 500 mg po 0.8 0.3 no Herbert 2002 [52]

Levofloxacin human ph non-inf 16 750 mg po 2.8 0.5 partially George 2010 [38]

Moxifloxacin human ph non-inf 13 2x400 mg po bid 1.3 after 2 doses 0.4 yes Hariprasad 2006 [53]

Moxifloxacin human ph/ps non-inf 8 2x400 mg po bid 1.5 after 2 doses yes Fuller 2007 [54]

Moxifloxacin human ph/ps non-inf 21 1x400 mg po 0.6 yes Lott 2008 [55]

Moxifloxacin human ph/ps non-inf 27 1x400 mg po 0.1 0.1 no Vedantham 2006 [56]

Trimethoprim human ph non-inf 10 160 mg po bid 1.8 after 3 doses 0.40 partially Feiz 2013 [60]

Sulfamethoxazole human ph non-inf 10 800 mg po bid 5.9 after 3 doses 0.15 Feiz 2013 [60]

Clarithromycin human ph non-inf 21 500 mg po bid 0.3 after 6 doses 0.2 no Al-Sibai 1998 [61]

a) ph= phakic, ps = pseudophakic, aph = aphakic, a-v = aphakic-vitrectomized

b) inf = inflamed, non-inf = non-inflamed

c) refers to MICs of wild-type pathogens covered by the specific antibiotic agent

d) coverage dependent on species (see text for details)

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

(!) exceeding human doses

V/S = vitreous/serum antibiotic concentration ratio

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

table 2. EUCAST MIC breakpoints (susceptible ≤) of highly prevalent or otherwise significant organisms in bacterial endophthalmitis (ug/ml).

If breakpoints are not defined, ECOFFs (epidemiological cut-off values) are given if reasonable (*).

antibiotic Str. group Str. Str. viridans coagulase-neg. S.aureus Enterococci Haemophilus Entero- Pseudomonas observed maximum

A/B/C/G pneumoniae group staphylococci influenzae bacteriaceae aeruginosa levels (ug/ml)

Vancomycin 2 2 2 4 2 4 NA NA NA 5.4

Benzylpenicillin 0.25 0.06 0.25 [0.125] [0.125] NA NA NA NA

Ampicillin, Amoxicillina)

0.5 0.5a) a)

4 1 [8] NA 0.1 (- 1.6)

Piperacillin-Tazobactama) a) a) b) b) c) c)

8 16 0

Cefazoline 1 a)

NA 0.5b) b)

NA NA NA NA 3.0-6.7

Ceftriaxonea)

0.5 0.5b)

8* NA 0.125 1 NA 5.1

Ceftazidime NA NA NA NA NA NA NA 1 8 0-35.4

Cefepimea)

1 0.5b)

8* NA 0.25 1 8 2.9

Imipenema)

2 2 0.125* 0.125* 4 2 2 4 1.1-2.5

Meropenema)

2 2 0.5* 0.5* NA 2 2 2 8.9

Rifampicin 0.06 NA 0.06 0.06 NA NA NA NA 15.2

Linezolid 2 2 NA 4 4 4 NA NA NA 1.2-3.7

Daptomycin 1 NA NA 1 1 4d)

NA NA NA 12.4

Amikacin NA NA NA 8 8 NA NA 8 8 8.5

Gentamycin NA NA NA 1 1 NA NA 2 4 1.8

Ciprofloxacin NA NA NA 1 1 [4] 0.06 0.25 0.5 0.1-0.5

Levofloxacin 2 2 NA 1 1 [4] 0.06 0.5 [1] 0.6-2.8

Moxifloxacin 0.5 0.5 NA 0.25 0.25 NA 0.125 0.25 NA 0.1-1.5

TMP-SMX 2

1 1 NA 2 2 [0.03] 0.5 2 NA 1.8

Clarithromycin 0.25 0.25 NA 1 1 NA NA NA NA 0.3

a) inferred from benzylpenicillin susceptibility

b) inferred from oxacillin/cefoxitin susceptibility

c) inferred from ampicillin susceptibility

d) CLSI breakpoint1 cefazoline non-species-related breakpoint 1ug/ml

2 breakpoints are expressed as TMP concentration with a 1:19 TMP-SMX ratio

* ECOFF given; no breakpoints defined

[ ] square brackets: unfavourable therapy due to resistance rate and/or insufficient evidence for efficacy

NA not applicable